Y Combinator’s Garry Tan wants U.S. open-weight AI labs to ‘distill’ frontier models, too

6 days ago 29

Tan argues that frontier models themselves trained on public human knowledge so access to capable AI should be "a form of public good."